2
Towards Autonomous and Auditable Medical Imaging Model Development
Researchers have developed AMID, an autonomous multi-agent framework designed to streamline the development and validation of medical imaging models.
2
STEC: Evidence Compression for Deep Search in Open-domain Multi-Hop QA
The STEC framework introduces evidence compression to help search-based language models resolve conflicting information across multiple search trajectories in multi-hop question answering.
2
ECG-LDC: A Hardware-Efficient Low-Dimensional Computing Framework for ECG Arrhythmia Classification
The authors present ECG-LDC, a low-dimensional computing framework designed for energy-efficient and accurate arrhythmia classification on resource-constrained wearable devices.
2
VoxENES 2026: Benchmarking Generalization of Speech Spoofing Detectors Against LLM-Era TTS and Voice Conversion
The VoxENES 2026 benchmark is introduced to improve the detection of synthetic speech generated by modern LLM-based text-to-speech and voice conversion systems.
2
Invariant Learning Dynamics of Transformers in Inductive Reasoning Tasks
A new theoretical framework explains the emergence of inductive reasoning capabilities in transformer models by analyzing their learning dynamics across various synthetic tasks.
2
SALT-GNN: Handling Dense Neighborhoods in Anti-Money Laundering Graphs via Statistics-Aware Attention
Researchers developed SALT-GNN, a graph neural network framework that uses statistics-aware attention to improve the detection of money laundering in dense transaction networks.
2
Cross-Layer Misalignment Detection in Agent Skills: A Progressive Loading-Aware Contrastive Learning Approach
Researchers have developed a contrastive learning method to identify discrepancies between the natural-language descriptions of AI agent skills and their actual execution-time behaviors.
2
MM-ToolSandBox: A Unified Framework for Evaluating Visual Tool-Calling Agents
A new evaluation framework called MM-ToolSandBox has been introduced to assess the performance of visually grounded AI agents across diverse tool-calling tasks.
2
Characterizing Large Language Model Agentic Workflows: A Study on N8n Ecosystem
Researchers conducted an empirical analysis of how large language model agents are utilized within low-code and no-code automation platforms.
2
A Knowledge-Based Multi-Agent Framework for Security Control Recommendation
Researchers have proposed a security decision support system that uses a multi-agent framework to recommend security controls based on minimal user requirements.
2
KGCQual: An Interpretable Framework for Evaluating the Knowledge Graph Construction Quality from Text
The KGCQual framework offers an interpretable metric to evaluate the structural and semantic quality of automatically constructed knowledge graphs.
2
Italian judge blocks Inwit's efforts to stop Telecom Italia, Fastweb+Vodafone tower agreement exits
An Italian court has ruled against Inwit in a dispute regarding telecommunications tower agreements, though the company intends to appeal.
2
GES-TSP: Graph Edge Sparsification for TSP
Researchers have proposed Graph Edge Sparsification, a machine learning-based approach designed to improve the computational efficiency of solving large-scale Traveling Salesman Problems.
2
Faithful, Not Corrective: Message-Format Effects in Multi-Hop Agent Relays Are Tier-Dependent
Researchers investigated how message formatting affects information fidelity and generation quality when LLM agents pass information across multiple hops.
2
PhenoEmbed: Self-Supervised Multispectral UAV Time-Series Embeddings for Individual Tree Crown Phenology
PhenoEmbed is a self-supervised temporal embedding model designed to track the changing characteristics of individual tree crowns using multispectral UAV time-series data.
2
Towards Predictive, Aligned, and Scalable Robot Learning
Researchers introduced Lumo-2, a latent world-action model designed to improve robot learning by reasoning over physical world dynamics.
2
Beyond Claude Code: the Chinese AI tools poised to benefit after back-door alert
A cybersecurity warning from Chinese authorities regarding a backdoor in Anthropic's Claude Code is expected to drive local developers toward domestic AI coding alternatives.
2
9点1氪丨“国产存储第一股”长鑫科技公布承销团阵容；SK海力士登陆美股，上市首日大涨近13%；OpenAI推出ChatGPT智能体
This news roundup highlights SK Hynix's successful debut on the Nasdaq, the upcoming IPO of Chinese memory manufacturer CXMT, and OpenAI's launch of ChatGPT agents.
2
智谱CEO唐杰发内部信：“GLM 时刻”和万亿俱乐部之后，什么是更重要的事
Zhipu AI founder Tang Jie announced in an internal letter that the company will prioritize long-term AGI goals, autonomous agents, and safety over short-term monetization.
2
Nested-ReFT: Efficient Reinforcement Learning for Large Language Model Fine-Tuning via Off-Policy Rollouts
The paper introduces Nested-ReFT, an efficient reinforcement learning method for fine-tuning large language models on complex reasoning tasks using off-policy rollouts.
2
Beyond Sally-Anne: Evaluating Theory of Mind in LLMs using Epistemic Schelling Points
Researchers propose a new dialogue-based evaluation framework to more accurately assess the Theory of Mind capabilities of large language models.
2
TENET: One Step Toward Test-Driven Development for Repository-Level Code Generation
The TENET framework aims to facilitate repository-level test-driven development by enabling AI agents to synthesize code based on developer-defined test specifications.
2
Multi-Agent Routing as Set-Valued Prediction: A WildChat Benchmark and Cost-Aware Evaluation
A new benchmark for multi-agent routing evaluates how effectively models select appropriate agents for tasks while managing execution costs.
2
The Paternalistic Filter: Epistemic Injustice and Differential Refusal in LLM-Mediated History Education for Marginalized Romanian Students
A study reveals that LLMs acting as history tutors exhibit epistemic paternalism and biased refusal patterns when interacting with different student demographics.